基于肺癌CT建立淋巴结转移的诺莫图预测模型研究

目的探讨基于CT影像建立的诺莫图模型在肺癌淋巴结转移预测中的作用。方法回顾性分析了2014-2017年辽宁省肿瘤医院收治的211例恶性肺结节患者的影像和临床资料,其中72例经病理证实存在淋巴结转移。通过提取和筛选肺CT影像组学特征,构建诺模图模型对淋巴结转移进行预测。通过绘制ROC曲线并计算AUC值评估模型的预测能力,使用决策曲线分析评估模型的临床适用性。结果构建的诺莫图模型在训练集和测试集上的AUC分别为0.859(灵敏度为0.810,特异度为0.773)和0.864(灵敏度为0.820,特异度为0.753),决策曲线表明模型有良好的临床应用价值。结论基于CT图像特征以及相关临床指标构建的诺莫图模型是作为无创预测恶性肺结节淋巴结转移的有效方法。

Study on Nomogram Prediction Model of Lymph Node Metastasis Based on Lung Cancer CT

Objective To evaluate the role of CT imaging-based nomogram model in prediction of lymph node metastasis in lung cancer.Methods The imaging and clinical data of 211 patients with malignant pulmonary nodules admitted to the Liaoning Provincial Cancer Hospital from 2014 to 2017 were retrospectively analyzed.A total of 72 cases were pathologically confirmed with lymph node metastasis.Through the extraction and screening of CT radiomics features of lung,a nomogram model was constructed to predict lymph node metastasis.The clinical applicability of the model was evaluated by decision curve analysis.The predictive ability of the algorithm was evaluated by plotting the ROC curve and calculating the AUC value.Results The AUC of the established nomogram model on the training and the test sets were 0.859(sensitivity of 0.810,specificity of 0.773)and 0.864(sensitivity of 0.820,specificity of 0.753).The decision curve indicated that the nomogram model had good clinical value.Conclusion The nomogram model,based on CT image features and related clinical indicators,has great potential in non-invasive prediction of lymph node metastasis in patients with malignant pulmonary nodules.